Apple M1 Microarchitecture Research by Dougall Johnson

Firestorm: Overview | Base Instructions | SIMD and FP Instructions
Icestorm:  Overview | Base Instructions | SIMD and FP Instructions

SSHLL2 (8H)

Test 1: uops

Code:

  sshll2 v0.8h, v0.16b, #3
  movi v0.16b, 1
  movi v1.16b, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 1.000

Integer unit issues: 0.000

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire (01)cycle (02)033f4e51inst issue (52)~issue fp/simd (54)~dispatch fp/simd (57)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op fp/simd (7e)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst neon or fp (9a)a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100420371661168625100010001000264521020182037203715713189510001000100020372037111001100001873116111786100020382038203820382038
10042037156116862510001000100026452102018203720371571318951000100010002037203711100110001073116111786100020382038203820382038
10042037166116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
10042037166116862510001000100026452102018203720371571318951000100010002037203711100110000373116111786100020382038203820382038
10042037166116862510001000100026452112018203720371571318951000100010002037203711100110001073116111786100020382038203820382038
10042037166116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
10042037156116862510001000100026452112018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
10042037166116862510001000100026452102018203720371571318951000100010002037203711100110000073116111786100020382038203820382038
10042037166116862510001000100026452112018203720371571318951000100010002037203711100110000973116111786100020382038203820382038
10042037156116862510001000100026452112018203720371571318951000100010002037203711100110000073116111786100020382038203820382038

Test 2: Latency 1->2

Code:

  sshll2 v0.8h, v0.16b, #3
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03080b1e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)72scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa6a8acc2c5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042003715000002051968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000100071011611197910100001002003820038200382003820038
102042003715500907261968625101001001000010010000500284752102005420037200371842103187971010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
102042003715000002101968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037211020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752112001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
10204200371500000611968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038
102042003715000001031968625101001001000010010000500284752102001820037200371842103187451010020010000200100002003720037111020110099100100100001000000071011611197910100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)031e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a6a7a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000006402162219786010000102003820038200382003820038
1002420037150007119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000006402162219786010000102003820038200382003820038
100242003715000611968625100101010000101000050284752112001820037200841844331876710010201000020100002003720037111002110910101000010000001476402162219786010000102003820038200382003820038
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100002006404162219786010000102003820038200382003820038
1002420037149006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000006402162219786010000102003820038200382003820038
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100001006402162219786010000102003820038200382003820038
10024200371500061196862510010101000010100005028475211200182003720037184433187671001020100002010000200372003711100211091010100001000000126402162219786010000102003820038200382003820038
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100000006402162219786010000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100001006402162219786010000102003820038200382003820038
1002420037150006119686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000100001006402162219786010000102003820038200382003820038

Test 3: throughput

Count: 8

Code:

  sshll2 v0.8h, v8.16b, #3
  sshll2 v1.8h, v8.16b, #3
  sshll2 v2.8h, v8.16b, #3
  sshll2 v3.8h, v8.16b, #3
  sshll2 v4.8h, v8.16b, #3
  sshll2 v5.8h, v8.16b, #3
  sshll2 v6.8h, v8.16b, #3
  sshll2 v7.8h, v8.16b, #3
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)030b18191e1f3a3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a1a6a7a8a9acc2c5branch mispredict (cb)cdcfd5d6inst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042004815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
80204200381500004380029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000000029258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039
802042003815000060086258010810080008100800205006401322001920038200389977699898012020080032200800322003820038118020110099100100800001000000000011151180160200350800001002003920039200392003920039

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)033f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa8acc2cfd0icache miss (d3)d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242005115039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010004502000216532003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820062118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010030502000216322003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000502000216232003580000102003920039200392003920039
800242003815039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400000200192003820038999631001880010208000020800002003820038118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000502000216222003580000102003920039200392003920039
800242003815039258001010800001080000506400001200192003820038999631001880010208000020800002003820038118002110910108000010000502000216322003580000102003920039200392003920039